{"thread":{"id":"34584","subject":"What's cooking in git.git (Aug 2013, #01; Thu, 1)","startedAt":"2013-08-01T21:44:12Z","lastAt":"2013-08-06T16:37:52Z","messageCount":6,"participants":["Junio C Hamano","Thomas Rast"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"224436","messageId":"7vbo5hqe6r.fsf@alter.siamese.dyndns.org","threadId":"34584","inReplyTo":null,"subject":"What's cooking in git.git (Aug 2013, #01; Thu, 1)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2013-08-01T21:44:12Z","receivedAt":"2013-08-01T21:44:12Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Here are the topics that have been cooking.  Commits prefixed with\n'-' are only in 'pu' (proposed updates) while commits prefixed with\n'+' are in 'next'.\n\nThe first release candidate v1.8.4-rc1 has been tagged; only\nregression fixes and l10n updates from now on.\n\nYou can find the changes described here in the integration branches\nof the repositories listed at\n\n    http://git-blame.blogspot.com/p/git-public-repositories.html\n\n--------------------------------------------------\n[Graduated to \"master\"]\n\n* jc/rm-submodule-error-message (2013-07-25) 1 commit\n + builtin/rm.c: consolidate error reporting for removing submodules\n\n\n* jx/clean-interactive (2013-07-24) 2 commits\n  (merged to 'next' on 2013-07-30 at b7fd474)\n + git-clean: implement partial matching for selection\n  (merged to 'next' on 2013-07-24 at 1e63bb9)\n + Documentation/git-clean: fix description for range\n\n\n* lf/echo-n-is-not-portable (2013-07-29) 1 commit\n  (merged to 'next' on 2013-07-30 at e223ceb)\n + Avoid using `echo -n` anywhere\n\n\n* ma/hg-to-git (2013-07-23) 1 commit\n  (merged to 'next' on 2013-07-30 at 1dddc06)\n + hg-to-git: --allow-empty-message in git commit\n\n\n* ms/subtree-install-fix (2013-07-30) 1 commit\n + contrib/subtree: Fix make install target\n\n\n* ob/typofixes (2013-07-29) 1 commit\n  (merged to 'next' on 2013-07-30 at ec2c897)\n + many small typofixes\n\n--------------------------------------------------\n[New Topics]\n\n* mb/docs-favor-en-us (2013-08-01) 1 commit\n - Provide some linguistic guidance for the documentation.\n\n Will merge to and cook in 'next'.\n\n\n* jc/parseopt-command-modes (2013-07-30) 2 commits\n - tag: use OPT_CMDMODE\n - parse-options: add OPT_CMDMODE()\n\n Many commands use --dashed-option as a operation mode selector\n (e.g. \"git tag --delete\") that the user can use at most one\n (e.g. \"git tag --delete --verify\" is a nonsense) and you cannot\n negate (e.g. \"git tag --no-delete\" is a nonsense).  Make it easier\n for users of parse_options() to enforce these restrictions.\n\n--------------------------------------------------\n[Stalled]\n\n* tf/gitweb-ss-tweak (2013-07-15) 4 commits\n - gitweb: make search help link less ugly\n - gitweb: omit the repository owner when it is unset\n - gitweb: vertically centre contents of page footer\n - gitweb: ensure OPML text fits inside its box\n\n Comments?\n\n\n* rj/read-default-config-in-show-ref-pack-refs (2013-06-17) 3 commits\n - ### DONTMERGE: needs better explanation on what config they need\n - pack-refs.c: Add missing call to git_config()\n - show-ref.c: Add missing call to git_config()\n\n The changes themselves are probably good, but it is unclear what\n basic setting needs to be read for which exact operation.\n\n Waiting for clarification.\n $gmane/228294\n\n\n* jh/shorten-refname (2013-05-07) 4 commits\n - t1514: refname shortening is done after dereferencing symbolic refs\n - shorten_unambiguous_ref(): Fix shortening refs/remotes/origin/HEAD to origin\n - t1514: Demonstrate failure to correctly shorten \"refs/remotes/origin/HEAD\"\n - t1514: Add tests of shortening refnames in strict/loose mode\n\n When remotes/origin/HEAD is not a symbolic ref, \"rev-parse\n --abbrev-ref remotes/origin/HEAD\" ought to show \"origin\", not\n \"origin/HEAD\", which is fixed with this series (if it is a symbolic\n ref that points at remotes/origin/something, then it should show\n \"origin/something\" and it already does).\n\n Expecting a reroll, as an early part of a larger series.\n $gmane/225137\n\n\n* jk/list-objects-sans-blobs (2013-06-06) 4 commits\n . archive: ignore blob objects when checking reachability\n . list-objects: optimize \"revs->blob_objects = 0\" case\n . upload-archive: restrict remote objects with reachability check\n . clear parsed flag when we free tree buffers\n\n Attempt to allow \"archive --remote=$there $arbitrary_sha1\" while\n keeping the reachability safety.\n\n Seems to break some tests in a trivial and obvious way.\n\n\n* mg/more-textconv (2013-05-10) 7 commits\n - grep: honor --textconv for the case rev:path\n - grep: allow to use textconv filters\n - t7008: demonstrate behavior of grep with textconv\n - cat-file: do not die on --textconv without textconv filters\n - show: honor --textconv for blobs\n - diff_opt: track whether flags have been set explicitly\n - t4030: demonstrate behavior of show with textconv\n\n Make \"git grep\" and \"git show\" pay attention to --textconv when\n dealing with blob objects.\n\n I thought this was pretty well designed and executed, but it seems\n there are some doubts on the list; kicked back to 'pu'.\n\n\n* jc/format-patch (2013-04-22) 2 commits\n - format-patch: --inline-single\n - format-patch: rename \"no_inline\" field\n\n A new option to send a single patch to the standard output to be\n appended at the bottom of a message.  I personally have no need for\n this, but it was easy enough to cobble together.  Tests, docs and\n stripping out more MIMEy stuff are left as exercises to interested\n parties.\n\n Not ready for inclusion.\n\n Will discard unless we hear from anybody who is interested in\n tying its loose ends.\n\n\n* jk/gitweb-utf8 (2013-04-08) 4 commits\n - gitweb: Fix broken blob action parameters on blob/commitdiff pages\n - gitweb: Don't append ';js=(0|1)' to external links\n - gitweb: Make feed title valid utf8\n - gitweb: Fix utf8 encoding for blob_plain, blobdiff_plain, commitdiff_plain, and patch\n\n Various fixes to gitweb.\n\n Drew Northup volunteered to take a look into this.\n $gmane/226216\n\n\n* jc/show-branch (2013-06-07) 5 commits\n - show-branch: use commit slab to represent bitflags of arbitrary width\n - show-branch.c: remove \"all_mask\"\n - show-branch.c: abstract out \"flags\" operation\n - show-branch.c: lift all_mask/all_revs to a global static\n - show-branch.c: update comment style\n\n Waiting for the final step to lift the hard-limit before sending it out.\n\n--------------------------------------------------\n[Cooking]\n\n* tr/log-full-diff-keep-true-parents (2013-08-01) 1 commit\n - log: use true parents for diff even when rewriting\n\n Output from \"git log --full-diff -- <pathspec>\" looked strange,\n because comparison was done with the previous ancestor that touched\n the specified <pathspec>, causing the patches for paths outside the\n pathspec to show more than the single commit has changed.\n\n I am not sure if that is necessarily a problem, though.  Output\n from \"git log --full-diff -2 -- <pathspec>\" without this change\n will be applicable to some codebase, but after this change that\n will no longer be true (you will get only tiny parts of the change\n that were made by the two commits in question, while missing all\n the other changes).\n\n\n* bc/unuse-packfile (2013-08-01) 2 commits\n - Don't close pack fd when free'ing pack windows\n - sha1_file: introduce close_one_pack() to close packs on fd pressure\n\n Will merge to and cook in 'next'.\n\n\n* da/darwin (2013-07-30) 1 commit\n  (merged to 'next' on 2013-08-01 at d151a94)\n + imap-send: use Apple's Security framework for base64 encoding\n\n Will cook in 'next'.\n\n\n* nd/sq-quote-buf (2013-07-30) 3 commits\n  (merged to 'next' on 2013-08-01 at dc7934a)\n + quote: remove sq_quote_print()\n + tar-tree: remove dependency on sq_quote_print()\n + for-each-ref, quote: convert *_quote_print -> *_quote_buf\n\n Code simplification as a preparatory step to something larger.\n\n Will cook in 'next'.\n\n\n* jc/url-match (2013-07-31) 6 commits\n - config: \"git config --get-urlmatch\" parses section.<url>.key\n - builtin/config: refactor collect_config()\n - config: parse http.<url>.<variable> using urlmatch\n - config: add generic callback wrapper to parse section.<url>.key\n - config: add helper to normalize and match URLs\n - http.c: fix parsing of http.sslCertPasswordProtected variable\n\n Reroll of km/http-curl-config-per-url topic.  Peff raises many good\n points about the tests for http.* variables.\n\n Waiting for the discussion to conclude, hopefully with a replacement test.\n\n\n* jl/submodule-mv (2013-07-30) 5 commits\n - rm: delete .gitmodules entry of submodules removed from the work tree\n - mv: update the path entry in .gitmodules for moved submodules\n - submodule.c: add .gitmodules staging helper functions\n - mv: move submodules using a gitfile\n - mv: move submodules together with their work trees\n (this branch uses nd/magic-pathspec.)\n\n \"git mv A B\" when moving a submodule A does \"the right thing\",\n inclusing relocating its working tree and adjusting the paths in\n the .gitmodules file.\n\n Waiting for a reroll.\n\n\n* nd/clone-connectivity-shortcut (2013-07-23) 1 commit\n  (merged to 'next' on 2013-07-30 at 8d16721)\n + smart http: use the same connectivity check on cloning\n\n Will cook in 'next'.\n\n\n* db/http-savecookies (2013-07-30) 1 commit\n  (merged to 'next' on 2013-07-30 at d2ad6e5)\n + http: add http.savecookies option to write out HTTP cookies\n\n Will cook in 'next'.\n\n\n* sb/fsck-opt-bool (2013-07-29) 1 commit\n  (merged to 'next' on 2013-07-30 at 8a9964c)\n + fsck: Replace deprecated OPT_BOOLEAN by OPT_BOOL\n\n Will cook in 'next'.\n\n\n* rr/feed-real-path-to-editor (2013-07-29) 1 commit\n  (merged to 'next' on 2013-07-30 at 72c9bcb)\n + editor: use canonicalized absolute path\n\n Will cook in 'next'.\n\n\n* rj/doc-rev-parse (2013-07-22) 2 commits\n  (merged to 'next' on 2013-07-22 at 8188667)\n + rev-parse(1): logically group options\n + rev-parse: remove restrictions on some options\n\n Will cook in 'next'.\n\n\n* jk/fast-import-empty-ls (2013-06-23) 4 commits\n  (merged to 'next' on 2013-07-22 at d26dab6)\n + fast-import: allow moving the root tree\n + fast-import: allow ls or filecopy of the root tree\n + fast-import: set valid mode on root tree in \"ls\"\n + t9300: document fast-import empty path issues\n\n Will cook in 'next'.\n\n\n* jc/diff-filter-negation (2013-07-19) 6 commits\n  (merged to 'next' on 2013-07-22 at 80d167d)\n + diff: deprecate -q option to diff-files\n + diff: allow lowercase letter to specify what change class to exclude\n + diff: reject unknown change class given to --diff-filter\n + diff: preparse --diff-filter string argument\n + diff: factor out match_filter()\n + diff: pass the whole diff_options to diffcore_apply_filter()\n (this branch is used by jc/hold-diff-remove-q-synonym-for-no-deletion.)\n\n Teach \"git diff --diff-filter\" to express \"I do not want to see\n these classes of changes\" more directly by listing only the\n unwanted ones in lowercase (e.g. \"--diff-filter=d\" will show\n everything but deletion) and deprecate \"diff-files -q\" which did\n the same thing as \"--diff-filter=d\".\n\n Will cook in 'next'.\n\n\n* nd/magic-pathspec (2013-07-23) 48 commits\n  (merged to 'next' on 2013-07-23 at 784c3a1)\n + rm: do not set a variable twice without intermediate reading.\n  (merged to 'next' on 2013-07-22 at 39f7194)\n + t6131 - skip tests if on case-insensitive file system\n + parse_pathspec: accept :(icase)path syntax\n + pathspec: support :(glob) syntax\n + pathspec: make --literal-pathspecs disable pathspec magic\n + pathspec: support :(literal) syntax for noglob pathspec\n + kill limit_pathspec_to_literal() as it's only used by parse_pathspec()\n + parse_pathspec: preserve prefix length via PATHSPEC_PREFIX_ORIGIN\n + parse_pathspec: make sure the prefix part is wildcard-free\n + rename field \"raw\" to \"_raw\" in struct pathspec\n + tree-diff: remove the use of pathspec's raw[] in follow-rename codepath\n + remove match_pathspec() in favor of match_pathspec_depth()\n + remove init_pathspec() in favor of parse_pathspec()\n + remove diff_tree_{setup,release}_paths\n + convert common_prefix() to use struct pathspec\n + convert add_files_to_cache to take struct pathspec\n + convert {read,fill}_directory to take struct pathspec\n + convert refresh_index to take struct pathspec\n + convert report_path_error to take struct pathspec\n + checkout: convert read_tree_some to take struct pathspec\n + convert unmerge_cache to take struct pathspec\n + convert run_add_interactive to use struct pathspec\n + convert read_cache_preload() to take struct pathspec\n + line-log: convert to use parse_pathspec\n + reset: convert to use parse_pathspec\n + add: convert to use parse_pathspec\n + check-ignore: convert to use parse_pathspec\n + archive: convert to use parse_pathspec\n + ls-files: convert to use parse_pathspec\n + rm: convert to use parse_pathspec\n + checkout: convert to use parse_pathspec\n + rerere: convert to use parse_pathspec\n + status: convert to use parse_pathspec\n + commit: convert to use parse_pathspec\n + clean: convert to use parse_pathspec\n + guard against new pathspec magic in pathspec matching code\n + parse_pathspec: support prefixing original patterns\n + parse_pathspec: support stripping/checking submodule paths\n + parse_pathspec: support stripping submodule trailing slashes\n + parse_pathspec: add special flag for max_depth feature\n + convert some get_pathspec() calls to parse_pathspec()\n + parse_pathspec: add PATHSPEC_PREFER_{CWD,FULL} flags\n + parse_pathspec: save original pathspec for reporting\n + add parse_pathspec() that converts cmdline args to struct pathspec\n + pathspec: add copy_pathspec\n + pathspec: i18n-ize error strings in pathspec parsing code\n + move struct pathspec and related functions to pathspec.[ch]\n + clean: remove unused variable \"seen\"\n (this branch is used by jl/submodule-mv.)\n\n Use \"struct pathspec\" interface in more places, instead of array of\n characters, the latter of which cannot express magic pathspecs\n (e.g. \":(icase)makefile\" that matches both Makefile and makefile).\n\n Will cook in 'next'.\n\n\n* jc/check-x-z (2013-07-11) 4 commits\n  (merged to 'next' on 2013-07-15 at 62751bc)\n + check-attr -z: a single -z should apply to both input and output\n + check-ignore -z: a single -z should apply to both input and output\n + check-attr: the name of the character is NUL, not NULL\n + check-ignore: the name of the character is NUL, not NULL\n\n \"git check-ignore -z\" applied the NUL termination to both its input\n (with --stdin) and its output, but \"git check-attr -z\" ignored the\n option on the output side.\n\n This is potentially a backward incompatible fix.  I am tempted to\n merge this to and keep it in 'next' for a while to see if anybody\n screams before deciding if we want to do anything to help existing\n users (there may be none).\n\n Will cook in 'next'.\n\n\n* ms/fetch-prune-configuration (2013-07-18) 1 commit\n  (merged to 'next' on 2013-07-21 at 7ffb690)\n + fetch: make --prune configurable\n\n Allow fetch.prune and remote.*.prune configuration variables to be set,\n and \"git fetch\" to behave as if \"--prune\" is given.\n\n \"git fetch\" that honors remote.*.prune is fine, but I wonder if we\n should somehow make \"git push\" aware of it as well.  Perhaps\n remote.*.prune should not be just a boolean, but a 4-way \"none\",\n \"push\", \"fetch\", \"both\"?\n\n Will cook in 'next'.\n\n\n* km/svn-1.8-serf-only (2013-07-18) 3 commits\n  (merged to 'next' on 2013-07-19 at ab8d034)\n + Git.pm: revert _temp_cache use of temp_is_locked\n  (merged to 'next' on 2013-07-12 at 70e0b95)\n + git-svn: allow git-svn fetching to work using serf\n + Git.pm: add new temp_is_locked function\n\n Subversion 1.8.0 that was recently released breaks older subversion\n clients coming over http/https in various ways.\n\n Will cook in 'next'.\n\n\n* tr/merge-recursive-index-only (2013-07-07) 3 commits\n - merge-recursive: -Xindex-only to leave worktree unchanged\n - merge-recursive: untangle double meaning of o->call_depth\n - merge-recursive: remove dead conditional in update_stages()\n\n Will hold, until we get any user.\n\n\n* jc/push-cas (2013-08-01) 7 commits\n  (merged to 'next' on 2013-08-01 at a7604ee)\n + t5540/5541: smart-http does not support \"--force-with-lease\"\n  (merged to 'next' on 2013-07-23 at 72371b0)\n + t5533: test \"push --force-with-lease\"\n + push --force-with-lease: tie it all together\n + push --force-with-lease: implement logic to populate old_sha1_expect[]\n + remote.c: add command line option parser for \"--force-with-lease\"\n + builtin/push.c: use OPT_BOOL, not OPT_BOOLEAN\n + cache.h: move remote/connect API out of it\n\n Allow a safer \"rewind of the remote tip\" push than blind \"--force\",\n by requiring that the overwritten remote ref to be unchanged since\n the new history to replace it was prepared.\n\n The machinery is more or less ready.  The \"--force\" option is again\n the big red button to override any safety, thanks to J6t's sanity\n (the original round allowed --lockref to defeat --force).\n\n The logic to choose the default implemented here is fragile\n (e.g. \"git fetch\" after seeing a failure will update the\n remote-tracking branch and will make the next \"push\" pass,\n defeating the safety pretty easily).  It is suitable only for the\n simplest workflows, and it may hurt users more than it helps them.\n\n Also, smart-http needs to be taught about this before we can move\n it forward.\n\n Will cook in 'next'.\n\n\n* jc/pull-training-wheel (2013-07-19) 1 commit\n  (merged to 'next' on 2013-07-22 at 748180e)\n + pull: require choice between rebase/merge on non-fast-forward pull\n\n Make \"git pull\" (without arguments that say what branch to\n integrate from where) refuse with \"it does not fast forward; choose\n between 'pull --merge' and 'pull --rebase'\".\n\n See\n\n http://thread.gmane.org/gmane.comp.version-control.git/225146/focus=225326\n http://thread.gmane.org/gmane.comp.version-control.git/225146/focus=229221\n\n for full discussion.\n\n Will cook in 'next'.\n\n\n* rj/cygwin-clarify-use-of-cheating-lstat (2013-07-18) 1 commit\n  (merged to 'next' on 2013-08-01 at 3ebfe7c)\n + cygwin: Remove the Win32 l/stat() implementation\n\n Cygwin port added a \"not quite correct but a lot faster and good\n enough for many lstat() calls that are only used to see if the\n working tree entity matches the index entry\" lstat() emulation some\n time ago, and it started biting us in places.  This removes it and\n uses the standard lstat() that comes with Cygwin.\n\n Recent topic that uses lstat on packed-refs file is broken when\n this cheating lstat is used, and this is a simplest fix that is\n also the cleanest direction to go in the long run.\n\n Will merge to 'master'.\n\n\n* jc/reflog-doc (2013-06-19) 1 commit\n - setup_reflog_action: document the rules for using GIT_REFLOG_ACTION\n\n Document rules to use GIT_REFLOG_ACTION variable in the scripted\n Porcelain.  git-rebase--interactive locally violates this, but it\n is a leaf user that does not call out to or dot-sources other\n scripts, so fixing it is not all that urgent.\n\n\n* jn/add-2.0-u-A-sans-pathspec (2013-04-26) 1 commit\n - git add: -u/-A now affects the entire working tree\n\n Will cook in 'next' until Git 2.0.\n\n\n* jc/core-checkstat-2.0 (2013-05-06) 1 commit\n - core.statinfo: remove as promised in Git 2.0\n\n Will cook in 'next' until Git 2.0.\n\n\n* jc/push-2.0-default-to-simple (2013-06-18) 1 commit\n - push: switch default from \"matching\" to \"simple\"\n\n Will cook in 'next' until Git 2.0.\n\n\n* jc/add-2.0-ignore-removal (2013-04-22) 1 commit\n - git add <pathspec>... defaults to \"-A\"\n\n Updated endgame for \"git add <pathspec>\" that defaults to \"--all\"\n aka \"--no-ignore-removal\".\n\n Will cook in 'next' until Git 2.0.\n\n\n* jc/hold-diff-remove-q-synonym-for-no-deletion (2013-07-19) 1 commit\n - diff: remove \"diff-files -q\" in a version of Git in a distant future\n (this branch uses jc/diff-filter-negation.)\n\n Will cook in 'next' until a distant future.\n\n--------------------------------------------------\n[Discarded]\n\n* jh/remote-hg-fetch-fix (2013-07-25) 2 commits\n  (merged to 'next' on 2013-07-25 at 33161ad)\n + Revert \"remotes-hg: bugfix for fetching non local remotes\"\n  (merged to 'next' on 2013-07-24 at 9c96641)\n + remotes-hg: bugfix for fetching non local remotes\n\n Reverted.\n\n Waiting for the final patch to replace, after discussion settles.\n\n\n* km/http-curl-config-per-url (2013-07-31) 7 commits\n  (merged to 'next' on 2013-07-31 at e88aad5)\n + Revert km/http-curl-config-per-url in its entirety\n  (merged to 'next' on 2013-07-25 at 531e8cb)\n + docs: update http.<url>.* options documentation\n  (merged to 'next' on 2013-07-23 at f84689b)\n + config: allow http.<url>.* any user matching\n + tests: add new test for the url_normalize function\n + config: improve support for http.<url>.* settings\n + config: add support for http.<url>.* settings\n + http.c: fix parsing of http.sslCertPasswordProtected variable\n\n Reverted to be replaced with a reroll (jc/url-match topic).\n"},{"id":"224437","messageId":"7v7gg5qe5g.fsf@alter.siamese.dyndns.org","threadId":"34584","inReplyTo":"7vbo5hqe6r.fsf@alter.siamese.dyndns.org","subject":"What's in \"What's cooking\"","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2013-08-01T21:44:59Z","receivedAt":"2013-08-01T21:44:59Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Here is a note to explain various sections and what the readers on\nthis list are expected to read from them in the periodic \"What's\ncooking\" summary.  This probably should be integrated in the periodic\n\"A note from the maintainer\" posting eventually.\n\nThe first part is mostly a boilerplate message, but the second\nparagraph of it often explains where in the development cycle we are.\nThe latest one (issue #1 of this month) said:\n\n    The first release candidate v1.8.4-rc1 has been tagged; only\n    regression fixes and l10n updates from now on.\n\nThe preamble is followed by various sections, each listing topics in\ndifferent \"doneness\".\n\nEach topic looks like this:\n\n  * nd/sq-quote-buf (2013-07-30) 3 commits\n    (merged to 'next' on 2013-08-01 at dc7934a)\n   + quote: remove sq_quote_print()\n   + tar-tree: remove dependency on sq_quote_print()\n   + for-each-ref, quote: convert *_quote_print -> *_quote_buf\n\n   Code simplification as a preparatory step to something larger.\n\n   Will cook in 'next'.\n\nA line that begins with a '*' has the name of a topic, the date of the\nlast activity on the topic and the number of changes in the topic.\nThen each commit on the topic is listed with prefix (\"+\", \"-\" or \".\").\nThe ones with \"+\" are already on 'next', and the dates when they were\nmerged to 'next' also appear on this list. This will give readers the\nrough idea of how long the topic has been cooking in 'next' (the\nbranch active bug hunters and early adopters are expected to be\nrunning) to measure the \"doneness\".\n\nThe list of commits on the topic is often followed by a few lines of\nfree-form text to summarize what the topic aims to achieve. This often\ngoes literally to the draft release notes when the topic graduates to\n'master'.  Also I annotate my short-term plan for the topic (\"Will\ncook in 'next' in the above example); it is very much appreciated if\nreaders raise timely objections to this plan (e.g. \"Don't merge it\nyet, it is broken --- see that discussion thread\").\n\nThe [New topics] section lists topics that were not in my tree when\nthe previous issue of \"What's cooking\" was sent.\n\nThe [Graduated to \"master\"] section lists topics that have been\nmerged to 'master' since the previous issue. Even if you are not an\nactive bug hunter, please exercise the features/fixes these topics\nbrings to the system, as breakages or regressions caused by these\nwill appear in the next release if they are not noticed and fixed.\n\nThe [Stalled] section lists topics whose discussion has stalled, or\nre-rolled patches have not been seen for a while after the topic\nreceived review comments. This section is listed here so that people\ninterested in the topics in it can help the owners of them to move\nthings forward.\n\nThe [Cooking] section lists the remainder of the topics. They are\nmaking normal progress and are no cause for worries in general.\nThose whose commits are all in \"next\" are often just simmering to\ngive time to active bug hunters find unexpected problem in them.\nthose whose commits are not in \"next\" yet are still going through\nreview/revise cycle until they become ready.\n"},{"id":"224506","messageId":"87a9kz3uy4.fsf@hexa.v.cablecom.net","threadId":"34584","inReplyTo":"7vbo5hqe6r.fsf@alter.siamese.dyndns.org","subject":"Re: What's cooking in git.git (Aug 2013, #01; Thu, 1)","fromName":"Thomas Rast","fromEmail":"trast@inf.ethz.ch","sentAt":"2013-08-03T10:55:15Z","receivedAt":"2013-08-03T10:55:15Z","isPatch":false,"sender":{"key":"tr@thomasrast.ch","avatar":"https://avatars.githubusercontent.com/u/153510?v=4"},"body":"Junio C Hamano <gitster@pobox.com> writes:\n\n> * tr/log-full-diff-keep-true-parents (2013-08-01) 1 commit\n>  - log: use true parents for diff even when rewriting\n>\n>  Output from \"git log --full-diff -- <pathspec>\" looked strange,\n>  because comparison was done with the previous ancestor that touched\n>  the specified <pathspec>, causing the patches for paths outside the\n>  pathspec to show more than the single commit has changed.\n>\n>  I am not sure if that is necessarily a problem, though.  Output\n>  from \"git log --full-diff -2 -- <pathspec>\" without this change\n>  will be applicable to some codebase, but after this change that\n>  will no longer be true (you will get only tiny parts of the change\n>  that were made by the two commits in question, while missing all\n>  the other changes).\n\nHmm.  Uwe's original complaint was that --stat -- as in \"what other\nthings are touched when we modify foo\" -- is nonsensical.\n\nIn addition, applying what you describe above would be a very strange\nform of rebase: one that squashes everything into the commits that\naffect the given files.  When would it ever make sense to do such\nsquashing?  After all, you'd lose all the commit splits&messages in\nbetween.\n\nIf you care about it, I can introduce a new flag that lets the user\npick; it's pretty trivial.  But it seems very strange to me.\n\n-- \nThomas Rast\ntrast@{inf,student}.ethz.ch\n"},{"id":"224578","messageId":"7vsiyonp2w.fsf@alter.siamese.dyndns.org","threadId":"34584","inReplyTo":"87a9kz3uy4.fsf@hexa.v.cablecom.net","subject":"Re: What's cooking in git.git (Aug 2013, #01; Thu, 1)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2013-08-05T15:18:31Z","receivedAt":"2013-08-05T15:18:31Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Thomas Rast <trast@inf.ethz.ch> writes:\n\n> Junio C Hamano <gitster@pobox.com> writes:\n>\n>> * tr/log-full-diff-keep-true-parents (2013-08-01) 1 commit\n>>  - log: use true parents for diff even when rewriting\n>>\n>>  Output from \"git log --full-diff -- <pathspec>\" looked strange,\n>>  because comparison was done with the previous ancestor that touched\n>>  the specified <pathspec>, causing the patches for paths outside the\n>>  pathspec to show more than the single commit has changed.\n>>\n>>  I am not sure if that is necessarily a problem, though.  Output\n>>  from \"git log --full-diff -2 -- <pathspec>\" without this change\n>>  will be applicable to some codebase, but after this change that\n>>  will no longer be true (you will get only tiny parts of the change\n>>  that were made by the two commits in question, while missing all\n>>  the other changes).\n>\n> Hmm.  Uwe's original complaint was that --stat -- as in \"what other\n> things are touched when we modify foo\" -- is nonsensical.\n>\n> In addition, applying what you describe above would be a very strange\n> form of rebase: one that squashes everything into the commits that\n> affect the given files.  When would it ever make sense to do such\n> squashing?  After all, you'd lose all the commit splits&messages in\n> between.\n\nI am not so worried about actually applying that kind of patch, but\nmore about reviewing what happened in each of the single logical\nstep shown.  If you made two changes in two far-apart commits to\nfiles you are interested in and want to look at the changes made to\nother files to make each of them a complete solution, depending on\nhow the change was split across files, you would get useful to\nuseless result with your patch.  In one extreme, if your commits are\ntoo fine-grained and you committed one path at a time, then\n\"full-diff\" will not show anything useful, as the commits that hit\nthe pathspec do not have changes to any other paths.  In another\nextreme, you may have preparatory changes to other paths in commits\nthat immediately precede the commit that hits the pathspec and then\nfinally conclude the topic by introducing a caller of the prepared\ncodepath in other files to the file you are interested in, and your\npatch will show only the endgame change without showing the\npreparatory steps, which may or may not be useful, depending on what\nyou are looking for.\n\nSo while I do understand why sometimes you wish to see full diff \n\"git log --full-diff -- <pathspec>\" to match output from \"git show\"\nwithout pathspec, I am not sure doing it unconditionally and by\ndefault like your patch does is the best way to go.\n\nIn the meantime:\n\n    git rev-list --header --parents HEAD -- <pathspec> |\n    git -p diff-tree -p --stdin\n\nwould be one way to do so.\n"},{"id":"224657","messageId":"87pptryyh0.fsf@linux-k42r.v.cablecom.net","threadId":"34584","inReplyTo":"7vsiyonp2w.fsf@alter.siamese.dyndns.org","subject":"Re: What's cooking in git.git (Aug 2013, #01; Thu, 1)","fromName":"Thomas Rast","fromEmail":"trast@inf.ethz.ch","sentAt":"2013-08-06T09:12:27Z","receivedAt":"2013-08-06T09:12:27Z","isPatch":false,"sender":{"key":"tr@thomasrast.ch","avatar":"https://avatars.githubusercontent.com/u/153510?v=4"},"body":"Junio C Hamano <gitster@pobox.com> writes:\n\n> Thomas Rast <trast@inf.ethz.ch> writes:\n>\n>> Junio C Hamano <gitster@pobox.com> writes:\n>>\n>>> * tr/log-full-diff-keep-true-parents (2013-08-01) 1 commit\n>>>  - log: use true parents for diff even when rewriting\n>>>\n>>>  Output from \"git log --full-diff -- <pathspec>\" looked strange,\n>>>  because comparison was done with the previous ancestor that touched\n>>>  the specified <pathspec>, causing the patches for paths outside the\n>>>  pathspec to show more than the single commit has changed.\n>>>\n>>>  I am not sure if that is necessarily a problem, though.  Output\n>>>  from \"git log --full-diff -2 -- <pathspec>\" without this change\n>>>  will be applicable to some codebase, but after this change that\n>>>  will no longer be true (you will get only tiny parts of the change\n>>>  that were made by the two commits in question, while missing all\n>>>  the other changes).\n>>\n>> Hmm.  Uwe's original complaint was that --stat -- as in \"what other\n>> things are touched when we modify foo\" -- is nonsensical.\n[...]\n> I am not so worried about actually applying that kind of patch, but\n> more about reviewing what happened in each of the single logical\n> step shown.  If you made two changes in two far-apart commits to\n> files you are interested in and want to look at the changes made to\n> other files to make each of them a complete solution, depending on\n> how the change was split across files, you would get useful to\n> useless result with your patch.  In one extreme, if your commits are\n> too fine-grained and you committed one path at a time, then\n> \"full-diff\" will not show anything useful, as the commits that hit\n> the pathspec do not have changes to any other paths.  In another\n> extreme, you may have preparatory changes to other paths in commits\n> that immediately precede the commit that hits the pathspec and then\n> finally conclude the topic by introducing a caller of the prepared\n> codepath in other files to the file you are interested in, and your\n> patch will show only the endgame change without showing the\n> preparatory steps, which may or may not be useful, depending on what\n> you are looking for.\n>\n> So while I do understand why sometimes you wish to see full diff \n> \"git log --full-diff -- <pathspec>\" to match output from \"git show\"\n> without pathspec, I am not sure doing it unconditionally and by\n> default like your patch does is the best way to go.\n\nI'm utterly unconvinced.\n\nFirst, note that the behavior only shows when you use an option that\nenables parent rewriting, such as --graph or --parents.\n\nSo if we went with a default-off option, there would be something like\n--diff-original-parents.  The user would have to discover this option,\nand use it whenever he wants to use --full-diff in combination with\n--graph.  To obtain the same diffs as without --graph.\n\nHow do you explain that to a user?  \"git-log has about 150 options.\nSome of them interact badly, but don't worry, there are options in there\nsomewhere that make the weirdness go away.\"\n\n  [No, I didn't actually count, but it seems a good estimate from\n  looking at\n\n    git grep -c -E '^-.*::' Documentation/rev-list-options.txt Documentation/diff-options.txt\n\n  ]\n\n\nSecond, gitk's option \"Limit diffs to listed paths\", corresponding to\nthe inverse of --full-diff, shows the commits as with git-show.  (Not\ninternally, but still, the format is very similar.)  In particular it\nonly shows the changes from the commit itself.\n\n\nThird, the only actual user data point I have is Uwe, who clearly\nexpected his diffs to remain the same across --graph.  There's a lot of\nselection bias in this, because a happy user would not have complained,\nbut we can look at his example again (I added a right-hand side to the\nrange to make it reproducible).\n\nWith the patched version:\n\n  $ git log --parents --stat v3.8..v3.9 --full-diff -- drivers/net/ethernet/freescale/fec.c\n  commit 8d7ed0f051caf192994c02fbefb859eac03d1646 14109a59caf93e6eae726a34bf2f0897508ec8c1\n  [...]\n      net: fec: fix regression in link change accounting\n  [...]\n   drivers/net/ethernet/freescale/fec.c | 1 +\n   1 file changed, 1 insertion(+)\n\nWith current master:\n\n  commit 8d7ed0f051caf192994c02fbefb859eac03d1646 14109a59caf93e6eae726a34bf2f0897508ec8c1\n  [...]\n      net: fec: fix regression in link change accounting\n  [...]\n   Documentation/sound/alsa/ALSA-Configuration.txt    |    5 +-\n   MAINTAINERS                                        |   22 +-\n   Makefile                                           |    2 +-\n   arch/alpha/Makefile                                |    2 +-\n  [...]\n   509 files changed, 9507 insertions(+), 7014 deletions(-)\n\n509 files.  I don't know about you, but I think that's completely\nuseless for reviewing anything.\n\n\nThe only compromise I can see flying is using something like\n--diff-parents=original and --diff-parents=rewritten, defaulting to\n'original'.  Behind the scenes they would toggle the appropriate\nmachinery: 'rewritten' would enable parent simplification even when it\nis not otherwise enabled, and 'original' would enable saving parents\nwhenever simplification is enabled.  Naturally all of this only applies\nwith --full-diff.\n\n\n> In the meantime:\n>\n>     git rev-list --header --parents HEAD -- <pathspec> |\n>     git -p diff-tree -p --stdin\n>\n> would be one way to do so.\n\nWell, as a data point on how realistic this is, consider two things:\n\n* I've hacked around on git for about five years now, and do not\n  remember ever noticing or using --header.  Nor, for that matter, ever\n  using rev-list|diff-tree for actual work.\n\n* It doesn't work at my end: it doesn't show any diffs.  Removing\n  --header helps, but then you lose the log messages.  --pretty again\n  breaks the diffs.\n\n-- \nThomas Rast\ntrast@{inf,student}.ethz.ch\n"},{"id":"224685","messageId":"7vob9ahj1b.fsf@alter.siamese.dyndns.org","threadId":"34584","inReplyTo":"87pptryyh0.fsf@linux-k42r.v.cablecom.net","subject":"Re: What's cooking in git.git (Aug 2013, #01; Thu, 1)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2013-08-06T16:37:52Z","receivedAt":"2013-08-06T16:37:52Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Thomas Rast <trast@inf.ethz.ch> writes:\n\n>> So while I do understand why sometimes you wish to see full diff \n>> \"git log --full-diff -- <pathspec>\" to match output from \"git show\"\n>> without pathspec, I am not sure doing it unconditionally and by\n>> default like your patch does is the best way to go.\n>\n> I'm utterly unconvinced.\n\nI was waiting you to say \"I'm positively convinced that showing\nlocal diff with the true parent is right!\"; the reason why I planted\nthat seed in \"What's cooking\" to discuss this was because I needed\nto play devil's advocate to make sure that all of us know that this\nchange is a good one, and that we know we agree that this change is\na good one, as the change is large insemantics, and the bookkeeping\nburden is not negligible.  I suspect that we may want to restrict\nthe copying of the true parents only to limited cases (e.g. when\n\"--full-diff\" is given), but I think that can be done on top, as it\nis only the performance thing, not correctness.\n"}]}